This research will develop and test a methodology for identifying regions of excess cancer burden for breast and colorectal cancer in Iowa. It will refine measures of geographic access to cancer prevention, treatment, and screening services in Iowa by computing values using fine-scaled geographic data on individuals, the spatial choices of individuals, and the locations of services providers, It will use Iowa Cancer Registry (SEER) data for a ten year period (1900-1999) and linked patient files to Medicare and selected medical insurance records. It will compute statistical models from the family of logistic regression models to associate specific cancer burden measures to predictor vanables that capture local characteristics of the area and characteristics of the individuals. The cancer burden measures for breast and colorectal cancer are stage at diagnosis, five-year survival, and probability of screening. The methodology will develop a regional simulation workbench (RSW) to generate the expected range and variations in the cancer burden measures for small geographic areas of Iowa based on local demographic characteristics of the area and statewide cancer burden rates. Stochastic variations will be computed using Monte Carlo simulation methods. Regions will be identified using modified, geographic feature extraction methods. The methodology will be validated in Iowa and, in year three, plans will be developed for its adoption and implementation in two other states with strong cancer registries belonging to the North American Association of Cancer Registries. Results can be used to plan more appropriate cancer prevention and control programs. The researchers are geographers who specialize in geographic information science and medical geography and epidemiologists with a substantial record of research in cancer incidence, prevention and control.